Reply: De novo SPTAN1 mutation in axonal sensorimotor neuropathy and developmental disorder

We read with great interest the Letter to the Editor of Ylikallio et al. (2020) in response to our recent publication that describes SPTAN1 nonsense mutations associated with hereditary motor neuropathy (Beijer et al., 2019). Previously, mutations in SPTAN1 were associated with a range of epileptic and intellectual disability phenotypes. Ylikallio et al. (2020) present an interesting patient harbouring a de novo c.6367del (p.Val2123Cysfs*45) change in SPTAN1 (RefSeq NM_001130438.3) associated with early-onset sensorimotor neuropathy as well as neuropsychological deficiencies, therefore indicating an intriguing overlap between the phenotypes described by us and the earlier described neurodevelopmental phenotypes. The peripheral nervous system involvement seems to also encompass the sensory nerves; previously, only motor predominant phenotypes had been linked with SPTAN1 mutations. Also, the authors provide the first report of a de novo SPTAN1 mutation as the cause of peripheral neuropathy. In neurodevelopmental SPTAN1 phenotypes and indeed in neurodevelopmental disorders in general (in particular intellectual disability) the de novo occurrence of dominant mutations is a common mechanism of disease (Vissers et al., 2010).

The spectrin complex, formed through heterotetramerization of alpha-II-spectrin (SPTAN1) with one of four non-erythrocytic beta-spectrins (SPTBN1, SPTBN2, SPTBN4, SPTBN5) is of critical importance during neuronal development and homeostasis. The composition of the spectrin complex in terms of its beta-spectrin depends on the subcellular localization, with e.g. the alpha-II/beta-II-spectrin complex primarily localizing to the distal axon, where it interconnects actin rings in the axonal submembrane network. Conversely, in the somatodendritic area, the alpha-II/beta-III-spectrin complex is more prominent. While previously less well-defined, it was recently shown that the actin lattice in the somatodendritic area changes dynamically depending on the level of neuronal activity, possibly explaining the difficulty to obtain detailed insights so far concerning the spectrin complex in that subcellular compartment (Lavoie-Cardinal et al., 2020).

The crucial function of SPTAN1 is demonstrated by early lethality in several knockout animal models: Drosophila, Caenorhabditis elegans, and mice (Lee et al., 1993; Norman and Moerman, 2002). The mouse model generated by Huang et al. (2017) with cre-lox knockout of alpha-II spectrin specifically in peripheral sensory neurons, circumvents the lethality and corroborates the sensory deficiencies seen in the patient presented by Ylikallio et al. (2020). This model showed that the alpha-II spectrin-dependent cytoskeleton is dispensable for maintenance of small-diameter axons but essential to maintain axonal integrity of large-diameter myelinated axons, with sensory neurons lacking alpha-II spectrin showing a 50% decrease in conduction velocity and amplitude (Huang et al., 2017). Although a knockout cannot (fully) captivate a suspected haploinsufficiency pathomechanism, the symptoms seen in the mouse model raise the question why other patients carrying nonsense mutations show no signs of affected sensory neurons. In light of the motor neuropathy phenotype in several SPTAN1 patients, it is worth investigating whether similar defects in axonal integrity arise upon loss of SPTAN1 in peripheral motor neurons.

Including this latest publication, 52 patients with 30 different variants in SPTAN1 have been reported to date. With phenotypes ranging from early-infantile epileptic encephalopathy to late-onset hereditary motor neuropathy, with intellectual disability, epilepsy and hereditary spastic paraplegia phenotypes in between, alterations in SPTAN1 give rise to a truly broad phenotypic spectrum affecting either the central or peripheral nervous system or sometimes both (Syrbe et al., 2017; Gartner et al., 2018; Leveille et al., 2019). Mutations in SPTBN2 are associated with spinocerebellar ataxia (both SCA5 and SCA14) (Ikeda et al., 2006; Lise et al., 2012; Elsayed et al., 2014). In contrast, mutations in SPTBN4, both homozygous and compound heterozygous, are associated with axonal motor neuropathies which can be accompanied by intellectual disability, general hypotonia and deafness, emphasizing the vital role of the various components of the spectrin cytoskeleton in both the central and peripheral nervous system and mirroring, to an extent, the disease spectrum of SPTAN1 (Knierim et al., 2017; Hausler et al., 2020). Ylikallio et al. (2020) report the first patient with both peripheral neuropathy as well as neurodevelopmental delay, which allows us to theorize a phenotypical spectrum for SPTAN1 mutations. These contours of the expanding clinical spectrum are now steadily emerging with patients showing key differences to previous reports (Syrbe et al., 2017): in terms of symptoms present, such as the absence of seizures in an otherwise similar neurodevelopmental phenotype, as first reported by Gartner et al. (2018), an isolated peripheral motor neuropathy phenotype in our cases (Beijer et al., 2019), a distinct combined central and peripheral nervous system phenotype in the spastic paraplegia cases presented by Leveille et al. (2019) and now a neurodevelopmental phenotype with distinct peripheral sensorimotor neuropathy, reported by Ylikallio et al. (2020).

Currently, very little is known about the underlying possibly divergent pathomechanisms that drive the vast phenotypical heterogeneity associated with SPTAN1 variants. Similarly, this phenotypic spectrum is also represented in the group of beta-spectrin genes, as mutations in the different beta-spectrins mirror this neurological disease spectrum. We wonder whether the clue to unravelling SPTAN1’s phenotypical spectrum can be found by considering differences in alpha-II-spectrin’s obligatory, but variable, binding partners, especially since the composition of the spectrin complex differs depending on the subcellular localization and the neuronal subtype.

Data availability

Data availability is not applicable to this article as no new data were created or analysed in this study.

Competing interests

The authors report no competing interests.

References